Image sensors in digital cameras and the like often adopt an autofocus (“AF” below) method referred to as the TV-AF method. This method moves the focus position within a certain range and calculates distance to the subject using an AF evaluation signal at each of a number of points within this range (e.g., see the specification of Japanese Patent Application Laid-Open no. 3-68280). The AF evaluation signal is computed by extracting a signal component of a prescribed band using a BPF (bandpass filter), etc. The more the subject is in focus, the greater the signal computed.
For example, if the distance from infinity to 50 cm is adopted as the range of distance measurement, as shown in FIG. 10, an AF evaluation signal at a focus position for focusing on infinity is acquired and then AF evaluation signals at each of a number of distances are acquired successively while making the focus position approach to a distance of 50 cm. AF control is subsequently carried out through a method of comparing the AF evaluation signals acquired at each of the distances and taking the focus position to be at a subject distance A judged to provide the best focus. (The subject distance A corresponds to the peak of the AF evaluation signal.)
It is usually difficult to acquire AF evaluation signals continuously while assigning focus positions. In many cases, therefore, AF evaluation signals are acquired upon downsampling AF evaluation signals at intervals of distances equivalent to depth of field.
However, in the case of a digital camera having a long focal length and a broad range of focus movement or a digital camera having a shallow depth of field and requiring fine movement of focus, the prior art is such that it is necessary to increase greatly the number of data samples (AF evaluation signals) acquired with regard to the range of distance measurement. As a consequence, acquiring the data takes time and, hence, the time needed for AF control is prolonged.